1. Introduction
In recent years there have been great advances in our understanding of the mainly Proterozoic
hypabyssal dyke swarms intruded into Precam- brian cratons. Much information has come from
studies of mainly Phanerozoic continental flood basalt CFB provinces which have similar chem-
istry. Although field evidence of such a relation- ship is often lacking see arguments expressed in
Ross, 1983; Tarney, 1992; Cadman et al., 1995, it can sometimes be deduced that flood basalts were
fed from the extensive dyke systems which are now exposed in Precambrian cratons e.g. Baragar
et al., 1996.
Despite very detailed research on both phenom- ena, the degree to which various petrogenetic
processes such as fractional crystallisation, crustal contamination and mantle metasomatism control
their chemistry is still hotly debated. An obvious difficulty in any petrogenetic analysis of dykes is
that the same process may have operated on the magma at different stages in its genesis; for exam-
ple, crystal fractionation within a basaltic magma may take place both prior to dyke injection e.g.
within a magma chamber and subsequently within the dykes themselves, the relative influence
of fractionation within each environment may be very difficult to ascertain. Hence although many
studies of intradyke petrogenetic processes have been undertaken e.g. Gibb, 1968; Komar, 1972,
1976; Ross, 1983, 1986; Platten and Watterson, 1987; Blichert-Toft et al., 1992; Ernst and Bell,
1992 the degree to which basalt petrogenesis may be controlled by hypabyssal processes within
mafic dykes is still uncertain.
Study of the Kangaˆmiut dyke swarm offers an excellent opportunity to help resolve some of
these questions. Earlier workers have noted that unlike the vast majority of continental mafic
swarms, the dykes were injected into an overall contractional environment e.g. Escher et al.,
1976 and throughout much of their extent horn- blende is the dominant primary ferromagnesian
mineral Korstga˚rd, 1979; Bridgwater et al., 1995. However, the major element chemistry of
the swarm suggests a normal tholeiitic Fe-enrich- ment trend Escher et al., 1975; Bridgwater et al.,
1995. As hornblende is associated with calc – al- kaline fractionation, it would appear that the
petrogenetic processes governing the chemistry of the dykes may be unrelated to the crystallisation
processes within the dykes themselves. The field setting and unusual petrographical characteristics
of the Kangaˆmiut dyke swarm also require that models developed to explain the petrogenesis of
other dyke swarms are applied in order to test their validity. In this paper we seek to undertake
comprehensive major and trace element modelling of the chemistry of the Kangaˆmiut dykes with a
view to understanding the processes governing their formation.
2. Field relationships and geological setting
The Kangaˆmiut dyke swarm was emplaced into the high-grade gneisses of the Archaean craton of
SW Greenland Fig. 1, occurring in an area spanning 200 km south of and 100 km to the
north of the Nagssugtoqidian orogenic boundary Bridgwater et al., 1995. Although originally re-
garded as an ensialic orogen, many more recent studies of this orogeny suggest that it took the
form of a continent – continent collision between 2.1 and 1.7 Ga Kalsbeek et al., 1987; Marker et
al., 1995; Kalsbeek and Nutman, 1996a,b; Kriegs- man et al., 1996. Contacts between the dykes and
country rocks are usually sharp with little evi- dence of crustal remelting or absorption at the
margins of the dykes. The host rocks for the dyke swarm are mainly tonalitic and granodioritic
gneisses. Additionally, there is a tendency for paragneiss to be associated with zones of high
strain see Fig. 1. These paragneisses are gener- ally intermediate to highly siliceous in composi-
tion and sulphide-rich. Smaller quantities of marbles are also present. The protoliths of the
paragneisses are uncertain, but based on our field observation and compositional character, proba-
bly consisted of volcano-sedimentary sequences interbedded with small amounts of limestone.
Early descriptions of the Kangaˆmiut dykes were published by Ramberg 1948, Noe Nygaard
1952 and Escher et al. 1975. An extension of the dyke swarm is believed to occur in SE Green-
land e.g. Bridgwater et al., 1990. Within SW Greenland, the study of fault movement history
shows that the dykes are demonstrably younger than the 2.2 Ga high magnesian and tholeiitic
‘MD’ [‘m6etad6olerite’] dykes which intrude the southern part of the craton Berthelsen and Bridg-
water, 1960; Hall and Hughes, 1990. U – Pb zir- con SHRIMP analysis dated two the Kangaˆmiut
dykes at 2046 9 8 Ma and 2035 9 5 Ma Kalsbeek and Nutman, 1996a,b; Nutman et al., 1999. Re-
cent
40
Ar
39
Ar data confirm this 2040 Ma age Willigers et al., 1999.
Structural studies by Escher et al. 1976 showed that dyke orientation veered from NNE –
SSW to NE – SW approaching the Nagssugtoqid- ian orogenic boundary Fig. 1, with a second
subordinate set of ESE-trending intrusions also being present. Escher et al. 1976 suggested that
the two sets were coeval, and that they were intruded along conjugate shear fractures. How-
ever, later studies of field and cross-cutting rela- tionships suggest three distinct crosscutting sets,
the E – WESE – WNW set is oldest, being crosscut by the NNE – SSW and subsequently the NE – SW
Fig. 1. Geological Map of West Greenland. Simplified from Marker et al. 1995. Insets: a Map of Greenland showing location of main map area: black = Nagsugtoquidian belt; b Details of sample localities.
dykes Mengel et al., 1996. The NE – SW set constitute the main set of intrusions and include
the two dykes dated by Nutman and Kalsbeek 1996 at ca. 2.04 Ga. North of Itilleq Fjord,
Kangaˆmiut dykes were emplaced into Archaean E – W trending amphibolite facies zones.
3. Mineralogy and petrology